Research On Short Term Wind Power Forecasting With Wind Farm And Optimal Dispatching Of Grid-connected Wind Power

With the depletion of traditional energy and the continuous deterioration of the natural environment,renewable energy represented by wind energy has gradually attracted the attention of people.The development of wind power generation is of great significance for improving the energy structure and coping with climate change and energy crisis.However,wind energy has the characteristics of volatility and intermittency.Large-scale wind power grid connection will pose a huge threat to the reliable operation and economic dispatch of power systems.Improving the forecasting accuracy of wind power and establishing a reasonable power system scheduling strategy under wind power grid connection can improve the utilization rate of wind power and reduce the impact of wind power grid connection on the power system.In this study,the research on short-term wind power forecasting and economic dispatch with wind power has been conducted.The specific research contents are as follows:Firstly,based on the traditional dragonfly algorithm(DA),adaptive learning factors and differential evolution strategies are introduced,and then an improved dragonfly algorithm(IDA)is proposed.The improved algorithm is tested and analyzed by unimodal function and multimodal function.The results show that the solving accuracy and optimization stability of IDA are superior to other algorithms.Secondly,the support vector machine(SVM)is introduced to establish short-term prediction models for wind power.In view of the fact that the parameters of SVM have a great impact on the prediction performance of model,the improved dragonfly algorithm is adopted to select the penalty factor and nuclear parameters,and then the IDA-SVM wind power prediction model is established.The dataset of wind farm in autumn and winter is used as the experimental data set.Taking the wind speed of the wind farm,the sine and cosine values of the wind direction as the input,and the corresponding wind power as the output,the wind power wind farm in the next 48 hours is predicted.Comparing the prediction results with the results of other prediction models,the results show that IDA-SVM has higher prediction accuracy and stability,indicating that the proposed model is a more reliable shortterm wind power prediction model.Finally,the economic dispatch of wind power grid-connected power system is studied.The static economic dispatch model of wind farm incorporated power system is established with the minimum sum of the cost of thermal power generation and wind power generation as the goal.The penalty function method is used to deal with the equality constraints.After considering the uncertainties of wind power,the wind power is allocated according to a fixed proportion of load demand.Then an improved crow search algorithm(ICSA)combining Gaussian operator is proposed and applied to solve two classic scheduling cases.The calculation results of crow search algorithm(CSA)and particle swarm algorithm are compared.The experimental results show that the ICSA has good optimization characteristics and stability,which can effectively reduce system operating costs and achieve more economical scheduling.